def __init_

def init()

in lib/models/patchnet_pct.py [0:0]
49 lines of code
5 McCabe index (conditional complexity)

    def __init__(self, cfg, num_heading_bin, num_size_cluster, mean_size_arr):
        super().__init__()
        self.cfg = cfg
        self.num_heading_bin = num_heading_bin
        self.num_size_cluster = num_size_cluster
        self.mean_size_arr = mean_size_arr

        # center estimation module
        self.center_reg_backbone1 = PlainNet(input_channels=3, layer_cfg=[128, 128, 256], kernal_size=1,dilated=cfg['dilated'])
        self.center_reg_head1 = nn.Sequential(nn.Linear(259, 256), nn.BatchNorm1d(256), nn.ReLU(inplace=True), nn.Dropout(0.5),
                                             nn.Linear(256, 128), nn.BatchNorm1d(128), nn.ReLU(inplace=True),
                                             nn.Linear(128, 3))
        self.center_reg_un1 = nn.Sequential(nn.Linear(259, 256), nn.BatchNorm1d(256), nn.ReLU(inplace=True), nn.Dropout(0.5),
                                             nn.Linear(256, 128), nn.BatchNorm1d(128), nn.ReLU(inplace=True),
                                             nn.Linear(128, 1))

        self.center_reg_backbone2 = PlainNet(input_channels=3, layer_cfg=[128, 128, 256], kernal_size=1,dilated=cfg['dilated'])
        self.center_reg_head2 = nn.Sequential(nn.Linear(259, 256), nn.BatchNorm1d(256), nn.ReLU(inplace=True), nn.Dropout(0.5),
                                             nn.Linear(256, 128), nn.BatchNorm1d(128), nn.ReLU(inplace=True),
                                             nn.Linear(128, 3))
        self.center_reg_un2 = nn.Sequential(nn.Linear(259, 256), nn.BatchNorm1d(256), nn.ReLU(inplace=True), nn.Dropout(0.5),
                                             nn.Linear(256, 128), nn.BatchNorm1d(128), nn.ReLU(inplace=True),
                                             nn.Linear(128, 1))

        self.center_reg_backbone3 = PlainNet(input_channels=3, layer_cfg=[128, 128, 256], kernal_size=1,dilated=cfg['dilated'])
        self.center_reg_head3 = nn.Sequential(nn.Linear(259, 256), nn.BatchNorm1d(256), nn.ReLU(inplace=True), nn.Dropout(0.5),
                                             nn.Linear(256, 128), nn.BatchNorm1d(128), nn.ReLU(inplace=True),
                                             nn.Linear(128, 3))
        self.center_reg_un3 = nn.Sequential(nn.Linear(259, 256), nn.BatchNorm1d(256), nn.ReLU(inplace=True), nn.Dropout(0.5),
                                             nn.Linear(256, 128), nn.BatchNorm1d(128), nn.ReLU(inplace=True),
                                             nn.Linear(128, 1))
        self.feat_reg = nn.Sequential(nn.Conv2d(64, 64, kernel_size=1, stride=4), nn.BatchNorm2d(64), nn.ReLU(inplace=True),
                                      nn.Conv2d(64, 64, kernel_size=1, stride=4), nn.BatchNorm2d(64), nn.ReLU(inplace=True),
                                      nn.Conv2d(64, 64, kernel_size=1))
        # box estiamtion module
        assert cfg['backbone'] in ['plainnet', 'resnet', 'resnext', 'senet']
        if cfg['backbone'] == 'plainnet':
            self.box_est_backbone = PlainNet(input_channels=3, layer_cfg=[128, 128, 256, 512], kernal_size=3, padding=1)
        if cfg['backbone'] == 'resnet':
            self.box_est_backbone = resnet()
        if cfg['backbone'] == 'senet':
            self.box_est_backbone = senet()
        if cfg['backbone'] == 'resnext':
            self.box_est_backbone = resnext()

        self.box_est_head1 = nn.Sequential(nn.Linear(515+64, 512), nn.BatchNorm1d(512), nn.ReLU(inplace=True), nn.Dropout(0.5),
                                          nn.Linear(512, 256), nn.BatchNorm1d(256), nn.ReLU(inplace=True),
                                          nn.Linear(256, 3 + self.num_heading_bin*2 + self.num_size_cluster*4))
        self.box_est_head2 = nn.Sequential(nn.Linear(515+64, 512), nn.BatchNorm1d(512), nn.ReLU(inplace=True), nn.Dropout(0.5),
                                          nn.Linear(512, 256), nn.BatchNorm1d(256), nn.ReLU(inplace=True),
                                          nn.Linear(256, 3 + self.num_heading_bin*2 + self.num_size_cluster*4))
        self.box_est_head3 = nn.Sequential(nn.Linear(515+64, 512), nn.BatchNorm1d(512), nn.ReLU(inplace=True), nn.Dropout(0.5),
                                          nn.Linear(512, 256), nn.BatchNorm1d(256), nn.ReLU(inplace=True),
                                          nn.Linear(256, 3 + self.num_heading_bin*2 + self.num_size_cluster*4))


        init_weights(self, self.cfg['init'])
def __init__()

def init()
